Erik Kusch
,
PhD Student
Department
of Biology
Section for Ec
oinformatics &
Biodiversity
Center for
Biodiversi
ty Dy
namics
in a Changing World (BIOCHANGE)
Aarhus University
26/03/2021
[Study
Group]
Bayesian
Statistics
w
ith the Rethinking
M
aterial
1
Usually best addre
ssed through
multile
vel
models (GLMMs)!
26/03/2021
[Study
Group]
Bayesian
Statistics
w
ith the Rethinking
M
aterial
2
Attempt to account for un
observed heterogeneity
.
26/03/2021
[Study
Group]
Bayesian
Statistics
w
ith the Rethinking
M
aterial
3
Hidden
State
Poisson
Likeli-
hood of
n
Hierarchy of predictor variab
les!
26/03/2021
[Study
Group]
Bayesian
Statistics
w
ith the Rethinking
M
aterial
4
Probabil
ity
of discrete outcomes is
subtraction of adjacent values.
Logistic
Function
26/03/2021
[Study
Group]
Bayesian
Statistics
w
ith the Rethinking
M
aterial
5
Log
-cumulativ
e
probabilities
Cumulativ
e
Probabilities
Posterior uncertainty allows for more
informative
identificatio
n
of intercepts becau
se
they give us mea
n
and
uncertainty
.
Do not
match ex
actly…
26/03/2021
[Study
Group]
Bayesian
Statistics
w
ith the Rethinking
M
aterial
6
Positive Param
eter
estimates
indicate
increase
in probability
Decreasing log-cum
ulative
odds for outcomes k-below the maxim
um
shifts probabi
lity
mass towards k.
26/03/2021
[Study
Group]
Bayesian
Statistics
w
ith the Rethinking
M
aterial
7
26/03/2021
[Study
Group]
Bayesian
Statistics
w
ith the Rethinking
M
aterial
8
Dirichlet is a distributio
n
for probabili
ty
distributions.
α
is the
proportion
of the total mass in the resultin
g distribution.